chapter 5

For a number of years, researchers of traditional peoples as well as representatives of indigenous groups and organizations voiced alarm about the rapid decay or extinction of slowly accumulated, locally-adapted ecological knowledge. However, most of these assertions were expressed in vague and generalized terms, supported by anecdotal or impressionistic observations, and were not backed up by hard evidence or precise information about the kinds, degrees, rates and causes of knowledge erosion. But during the last couple of decades a very substantial body of work relevant to this topic has been carried out, though it is often dispersed among different publications focused more on other issues, resulting in a relatively rich empirical database from which concrete observations of general and specific patterns can be made.

Based on our review of this literature, one of the main overall tendencies that stands out across a broad range of situations is the dynamic complexity of TEK in space and time. At the local level, such complexity is marked by widespread sharing alongside internal variation, simultaneous continuity and change. In a large number of studies, we find descriptions of the local corpus of TEK as consisting of a spectrum of categories, valuations and practices which range from portions that are widely shared by a large fraction of society members to portions that are specific to certain subgroups or even individuals. This characteristic reflects the dynamic state of the knowledge system in the sense that access and participation in particular subsets of it depend on active processes of communication, learning, experience and experimentation that vary from individual to individual. In a similar vein, certain elements are regarded to be deeply ancestral, passed down from generation to generation over long time periods, while other elements are recognized as being acquired more recently, such as within the lifetimes of current members. The derivation of the corpus from endogenous as well as exogenous sources highlights the inherently fluid and shifting nature of TEK, which may therefore be conceived as undergoing a continuous regenerative process encompassing a complex mix of replication, loss, addition, and transformation of constituent elements. The particular balance between the collective and the individual, and the old and the new, portions of the knowledge corpus varies considerably from site to site and to some extent provides a measure of its vitality and resilience. Viewed from a global perspective, we are confronted with a complex panorama of both convergent and divergent trends which are conditioned by multiple social and ecological factors. Precisely because our present picture of this reality is so complex and heterogeneous, a simplified and comparative indicator is needed.

5.1. TEK Erosion

The principal convergent process that can be identified from the existing data record is the widespread loss or erosion of TEK in many parts of the contemporary world. Solid empirical, often quantitative, evidence either consistent with or directly confirming the hypothesis of TEK erosion has been documented at numerous sites distributed among many different national, cultural, and eco-geographic contexts. Viewed at a national scale, we find clear indications of this trend occurring in the countries of: Indonesia, Malaysia, Brunei, Thailand, Vietnam, Papua New Guinea, Guinea-Bissau, South Africa, Mozambique, Sudan, Tanzania, Kenya, Madagascar, Canada, the U.S.A., Mexico, Guatemala, Honduras, Brazil, Bolivia, Peru, Colombia, Ecuador, Argentina, Venezuela, the Federated States of Micronesia, India, Nepal, Bhutan, and Italy. The biogeographic settings where this is happening encompass grasslands, deserts, woody savannas, tropical rain forests, swamp forests, temperate forests, boreal forests, circum-polar zones, high mountains, low mountains, coastal zones, coral atolls, and volcanic islands. The people being affected include indigenous ethnic groups, immigrant groups, mestizo groups, and multi-ethnic populations. The economic types run the gamut from farmers to hunters, collectors, trappers, foragers, fisherfolk, pastoralists, and mixed combinations of these. In sum, the great quantity and diversity of parallel case studies appear to provide empirical confirmation of the hypothesis of TEK erosion at a global scale.

5.2. Explanatory Variables of TEK Variation and Change

A wide variety of change indicator variables have been implicated as causing or correlating with the observed variations and changes in TEK levels. These include: age, gender roles, formal education, parental schooling, language shift, bilingualism, market involvement, imported technology, occupational focus, wealth, land availability, public economic assistance, sedentism, habitat degradation, useful species extinction, distance to forest or town, migration, travel, interethnic contact, availability of western medicines or health clinics, religious belief, and values change. However, the direction and strength of these effects on knowledge vary considerably across the different study sites and with respect to the specific domains or types of knowledge. Furthermore, those types of knowledge which appear to be most vulnerable and endangered in some localities are amazingly robust in other places.

5.2.1. Age

As noted above, age is the most common social variable used to evaluate the diachronic change or continuity of TEK. Significant differences in knowledge according to age group are usually interpreted as indicating change whereas the absence of such differences imply ongoing retention. The erosional process is often manifested locally as a marked gap in the numbers of taxa, uses, preparations, skills, ecological relationships, etc. known by older versus younger generations. Age alone does not explain why knowledge change takes place so it is frequently analyzed in combination with other social and environmental variables (see section 6.3.2). In several case studies reviewed here, no significant differences between older and younger people's knowledge were recorded but variable interpretations about whether knowledge decline had occurred or is occurring were made based on correlations with other dynamic variables or other available information. For example, local informants' statements to the effect that "younger people know and work less than people did in the past" seemed to suggest the possibility that significant changes had already taken place in a more distant earlier time period. Some of the authors acknowledge that this result may also be explained by limitations in the research design, such as the way that knowledge is defined and measured or the size and composition of the study population sample.

5.2.2. Gender

Gender is an important social variable linked to dynamic processes of change because knowledge is usually unevenly distributed in accordance with gender-specific work, family and ritual roles and spheres of social interaction, and because socioeconomic development often affects men and women differently. For example, in a number of studies women (especially older women) were found to possess greater knowledge of herbal remedies because they tend to be responsible for family and especially child health care. Among different rural communities of coastal Brazil, a few women are exceptionally knowledgeable about medicinal plants and therefore are considered key repositories of collective ethnomedical wisdom. Caniago and Siebert report that among the Ransa Dayak of Kalimantan, Indonesia women are more familiar with the medicinal plants found in cultivated and early successional areas while men know more about medicinal species found in primary forest. Elsewhere Voeks has observed that disturbance species are often sources of ethnomedical innovation (e.g. introduced species) due to loss of primary forests and acculturation. Ross and Medin demonstrated that Tzotztil Maya (Chiapas, Mexico) women were able to free list more utilitarian plant taxa than men because their residence and work habits were less affected by recent changes. The special role of women for the transmission of knowledge about wild food and medicinal plants is recognized in some societies.

5.2.3. Language shift

Local language has been highlighted as crucially important for TEK preservation because much information about the environment is encoded in the lexicon, grammar and discourse. Despite this importance, relatively few studies have attempted to measure the impact of language shift on TEK from a quantitative perspective. Where this has been done, the result has invariably been significant. Benz et al. showed that empirical knowledge about plant use is both more diverse and more evenly shared by the Huastec people of Mexico, who speak an indigenous language, in comparison with mestizo and Spanish-speaking indigenous populations in the Sierra de Manantlán, also in Mexico. The authors suggest that even though traditional knowledge about plants suffered a decline that accompanied loss of the indigenous language in Manantlán, a considerable amount of such knowledge is able to survive the recent modernization process where it plays an important role in subsistence. In more controlled investigations, Zent and Reyes-García et al. found that fluency in the national language was negatively correlated with ethnobotanical knowledge among the Piaroa (Venezuela) and Tsimane (Bolivia) groups respectively. Surveys of ethnobiological naming ability among different Native American populations still speaking the indigenous language as their first language and inhabiting a similar desert environment have produced contrasting results. Among the Tohono O'odham of the southwestern U.S., children could name only a small fraction of the common plant and animal species in their local environment that their grandparents could whereas among the Seri of Sonora, Mexico, youths registered no significant difference from elders in the lexical recognition of culturally salient and ecologically representative animal species. The differences may be attributable to the fact that O'odham youth no longer participate regularly in collecting expeditions and other subsistence activities and have little "hands-on" experience with their surrounding environment, due mainly to the commanding attention of school and television in their daily lives, while the Seri youth are still actively involved in the traditional activities. The fact that native ethnobiological names should be lost by indigenous mother tongue speakers seems to suggest that traditional environmental language is one of the linguistic domains that first suffers decay in a context of rapid cultural and ecological change and may be a prelude to more generalized language shift.

5.2.4. Formal education

The variable of formal education is noteworthy because it often has an effect on the informal learning of TEK but that effect is not uniform from place to place. Several studies reported school achievement or attendance as negatively impacting types of TEK acquisition. In other studies, by contrast, schooling is positively associated with higher knowledge or shows no correlation. Where the relationship is negative, it has been explained as due to the fact that time spent in school detracts from the time children can devote to subsistence and other traditional activities or leads to devaluation of traditional knowledge. Where the relationship is positive, it has been interpreted as the result of greater exposure to nonlocal in addition to local knowledge or greater opportunity to interact and share knowledge with others. While the contrasting results may simply reflect distinct local realities, we should also be aware that reliable comparison across the different case studies is also problematic due to the fact that the research designs differ considerably in terms of the specific domains or subtypes of TEK investigated, the population samples selected, the data elicitation methods used, the ways that knowledge is measured, and the statistical analyses applied. It should also be recognized that schooling interacts with other social change variables, such as residence, literacy, language fluency, occupation, contact with outsiders and personal experiences, and these inter-variable interactions need to be probed or controlled for to obtain a better understanding of how and where education influences TEK transmission.

5.2.5. Market Integration

Researchers have hypothesized that the transition from a subsistence to a market-based economy would have a negative impact on TEK because integration into external markets entails the specialized extraction or production of fewer goods, the substitution of local natural products and manufactures for imported ones, and the creation of greater socioeconomic heterogeneity, thus undermining the pooling of knowledge and resources. Contrary to this generic expectation, where this relationship has been systematically tested the results have been inconsistent. For the Tawahka Amerindians of Honduras, Godoy et al. observed that integration into the market through the sale of agricultural crops or wage labor was associated with less knowledge of plants and animals and their interactions, but integration into the market through the sale of timber and nontimber forest products was associated with higher knowledge of wildlife. Among Tsimane' Amerindians of lowland Bolivia, Reyes-García et al. found that distance between village and market town was correlated with higher agreement about plant uses, but after 50 km. the level of agreement declined. When village-to-town distance was controlled, other canonical market indicators such as cash income and material wealth showed no significant correlation with consensual knowledge. Vadez et al. examined the effects of markets on Tsimane' farming practices. Although farmers did intensify cash crop production, households and villages more integrated into the market planted more cassava and rice varieties, intercropped more, and put more crops in new fields than more subsistence-focused households.

Another set of studies has addressed the impact of markets on the perceived value of biodiversity for local groups. Among communities affected by the commercialization of timber and nontimber forest products, it has been observed that the market exerts a significant influence on species' quantitative importance values (i.e. more highly ranked species tend to have markets or commercial value). At the same time, such valuations differ according to the type and degree of dependence on market activities. Lawrence et al. compared the valuation of forest plant species among immigrant and indigenous communities in lowland Amazonian Peru. They found that immigrants nominated more species deemed to be valuable than did Indians but the former group placed higher value on commercial timber species while the latter group tend to value more the species used for food, nontimber trade products, and noncommercialized construction materials. In communities with greater access to markets, over-exploitation of the most valuable species and alternative livelihood options contribute to a decrease in the perceived value of the forest. Shanley and Rosa examined the valuation of local species by Caboclos in a forested area of the lower Amazon, Brazil characterized by the decline of nontimber forest product trade and the rise of commercial logging in recent decades. They found that young people and employees (and their families) of the logging companies place higher value on the commercial timber species while older and economically independent people tend to emphasize nontimber (i.e. technological, medicinal) species.

5.2.6. Western Medicines

Even though a majority of the world's population depends partially or entirely on traditional medicine for most of their health needs, western biomedical technology (i.e. pharmaceuticals, vaccinations, health clinics, paramedics, physicians, hospitals) has spread to nearly every corner of the world in the past several decades and constitutes a major agent of biological and sociocultural change. The influence of this factor on TEK change is felt most directly in the domain of ethnomedical knowledge. Some investigators regard medicinal plant knowledge to be especially sensitive and vulnerable to sociocultural change, often with deleterious consequences for the health status of the affected populations because of the costs and difficult access of modern medical care. While the data record demonstrates that ethnomedical knowledge and practice continues to be important for many rural communities, in a number of areas the rich aboriginal ethnopharmacopoea is nevertheless diminishing from the collective memory as a result of modernization impacts. Local manifestations of this degenerative trend include the intergenerational knowledge gap referred to above (section 5.2.1) as well as the rarity or disappearance of specialist healers in the community. The role played by western medicines and health clinics in TEK erosion goes beyond merely replacing traditional remedies as health care options since they are also linked with other factors of social and ecological change, including population growth, migration, settlement pattern, trade, religion, and values change. For example, among the Piaroa of the Venezuelan Amazon, the establishment of state-sponsored biomedical programs and services in frontier zones has helped to induce downriver migration, settlement nucleation, interethnic contact and acculturation in a large proportion of the population since the 1970's. This transition is in turn associated with the loss of traditional ethnomedical knowledge among younger members which also reinforces dependence on the healthcare provided by the government.

While the intergenerational gap in ethnomedical knowledge referred to above is the more common situation, some local groups display no significant differences between older and younger people and therefore it would appear that their knowledge is not undergoing noticeable decline. It is interesting to note, however, that the respective authors cited here invoke different contextual as well as methodological factors to explain the unexpected parity across age groups. The studies by Kristensen and Lykke were carried out among different communities of the Gourounsi people in the Sahel region of Burkina Faso. Even though the region is characterized by increasing demographic pressure, socioeconomic change and habitat alteration, they found that inventories of medicinal and other useful woody plants do not differ appreciably by age or gender, but do so by community. They wrote that this result may be partially explained by three factors: (1) floristic diversity is relatively low in the Sahel and thus non-specialist knowledge is limited and quickly learned, (2) questions on use-categories were relatively broad (e.g. the medicinal category was not disaggregated), and (3) the study area is poor and relatively isolated, few people have been to school. The study by Lozada et al. took place in a rural mestizo community in northwestern Patagonia, Argentina. They recognize that the indifference of knowledge distribution by age or sex departs from the pattern observed in other communities within the same region and attribute this distinctive result to: (a) the relative homogeneity of knowledge of wild plants in this population, (b) the possibility that significant innovations could have occurred in previous generations but the situation has since become more stable, and (c) the small sample size.

The dynamic status of ethnomedical knowledge is reflected as well in case studies suggesting the historical acquisition of nontraditional herbal remedies in substitution of native ones. For example, in some localities a large proportion of the local ethnopharmacopoea is made up of exotic species or species harvested from anthropogenically disturbed habitats (i.e. cultivated or secondary growth areas). In sum, we find different scenarios of ethnomedical TEK change, not all of which can be reduced to a simple unilineal process of loss over time.

5.2.7. Habitat Degradation

The loss of natural habitats or extinction of species due to human activities has been identified as an important cause of TEK variation and change in some studies. Rocha Silva and Andrade measured and compared the cultural significance indices of useful plants in several communities of northeastern Brazil. They recorded higher indices of introduced versus native plants in relation to degree of deforestation and proximity to urban areas, thus indicating that loss of knowledge about native species was caused in part by deforestation of the Atlantic forest ecosystem. Meanwhile among Caboclo communities who inhabit another region of northern Brazil that is suffering a similar process of deforestation, researchers recorded testimony of older informants who lamented the present day scarcity or disappearance of formerly highly valued plant species and game animals. Thus use decline in this area is mainly being caused not by loss of interest but rather by habitat conversion and local tree extinction. Among the Gourounsi of Burkina Faso, the use rank profiles of woody plants for villages in areas with much greater human and livestock activity grouped close together on ordination diagrams while villages located in less exploited areas formed another cluster group. This same division was repeated with regard to villagers' perceptions of declining trends of these species, thus confirming that use valuations are influenced by vegetation dynamics related to environmental conditions and human impact.

5.2.8. Migration

Immigrant populations are often considered to have less detailed, accurate and adaptive knowledge of their surrounding lands than long-resident indigenous groups. This assumption was put to a test by Atran and colleagues by comparing ecological knowledge, beliefs and practices for two immigrant groups (Q'eqchi' Maya, Ladino) and one native group (Itza' Maya) in the Petén forest of northern Guatemala. The results of their research confirm that the native Itza' do indeed exhibit much greater knowledge and awareness of ecological complexity involving animals, plants and people as well as use and management practices favoring forest regeneration in comparison to the immigrant groups. However, they also found that the Spanish-speaking Ladinos are closer to the native Itza' in thought and action than the Q'eqchi'. The researchers suppose that the main reasons that the Ladinos have adapted better to this environment are closer contact with the more knowledgeable Itza' and similarly diffuse patterns of social networking, which allow for more fluid flow and exchange of information across social boundaries. Meanwhile the more conservative and cooperatively-organized Q'eqchi maintain a cultural model of nature and resource management derived from their aboriginal highland heritage. This study demonstrates the variable capacity of immigrant groups to pick up new ecological knowledge rather quickly which is conditioned by patterns of social organization and communication.

The exchange and absorption of folk knowledge between groups in dynamic contact situations has also been explored by comparing the ethnobotanical inventories and use patterns of indigenous versus nonindigenous groups. Some nonindigenous or colonist groups display large inventories of useful plants that compare favorably with the inventories of indigenous groups occupying similar ecoregions. Campos and Ehringhaus compared the use of palms among two indigenous communities (Yawanawá and Kaxinawá) and two folk communities (rubber tappers and ribeirinhos) in southwestern Amazonia (Brazil) and, as expected, found that the indigenous groups know significantly more about palm uses than the folk groups. However, they also found that substantial percentages (20-30%) of the uses cited by the indigenous groups were considered to be nontraditional uses acquired through contact with the colonists.

5.2.9. Values Change

Several studies have pointed to changing values and beliefs as an important reason for TEK erosion or change. Among tribal and nontribal groups of Western Gats, India the knowledge and consumption of wild food plants is declining mainly because of the social stigma attached to these low-status food types. Both mothers and their children expressed shame at having other people see them collecting wild food plants even though they recognize these foods to be healthy and nutritious. Among the Dusun of northern Borneo, young people consciously avoid learning medicinal species because they believe that this domain of knowledge connects them to the primitive lifestyle of their parents that they are seeking to overcome. For the Manus of Papua New Guinea, one of the effects of religious missionization was to discredit the authority and prestige of native healers. This in turn is linked with the loss of specialized ethnomedical knowledge. In the case of the Lacandon Maya of southern Mexico, Ross found that younger adults had significantly less intricate knowledge of ecological complexity, expressed in terms of plant-animal interactions, than older adults. He traces this degradation to a clear shift in their social orientation as well as fundamental differences in their views of the spiritual and ceremonial composition of the world. Estomba et al. link the decline of native medicinal plant use to corresponding decline of the aboriginal medical cosmovision in a Mapuche community in northwestern Patagonia, Argentina, in noting the low citation of cultural syndrome ailments and the fact that an exotic plant species is the most commonly recognized cure for them.

5.3. TEK Persistence, Transfer and Creation

The preceding sections highlighted the pervasive erosion of TEK in many rural localities around the world and the principal drivers but also emphasized the complex and variable nature of this process and mentioned some examples where this trend is not clearcut. Adding to this complicated picture, there are a few important case studies which have registered the persistence, transfer or even addition of TEK in societies undergoing cultural, economic, and ecological changes. As mentioned earlier, Zarger and Stepp report finding no significant differences in plant naming ability by age among Tzeltal Mayan children in the community of Mahosik' between 1968 and 1999 despite the fact that numerous sociopolitical, economic and environmental changes had occurred. The authors explain this surprising result by noting that the basic subsistence system and routine activities of children have not changed very much in the 30 year interval between the two studies.

Kristensen and Lykke found no direct evidence that knowledge of medicinal and other useful woody plants is eroding among the Gourounsi group of the Sahel region of Burkina Faso despite the fact that the region is marked by increasing demographic pressure growth, socioeconomic change and habitat alteration. They attribute this result to the people's strong sense of pride in their traditions and their preference toward using natural resources instead of introduced exotic crops and industrial substitutes.

Byg and Balslev investigated processes and factors affecting people's knowledge and use of palms among indigenous and nonindigenous residents of the Andean foothills in southeastern Ecuador. This study produced no statistically significant indications that knowledge loss is occurring among the indigenous Shuar although the Shuar's own anecdotal perception is that erosion has taken place among the younger generation. Moreover, they found evidence of knowledge transfer from the Shuar to nonindigenous colonists and among the Shuar themselves. By contrast, they also identify village remoteness, marginality and lack of access to modern goods and services as factors that are associated with more knowledge.

Guest examined the knowledge of shrimp ecology in a fishing community on the northern coast of Ecuador that has been changing rapidly as a result of road construction and the subsequent influx of people and technology, integration with the national economy, and adoption of shrimp mariculture. He found that the local inhabitants had acquired in a relatively short amount of time new knowledge about shrimp which was not correlated with age, gender, education or years of residence in the community but instead was explained by the length of their experience in shrimp farming.

One of the most complete and sophisticated studies of the TEK erosion/change process up until now was carried out by Reyes-García and colleagues among Tsimane Amerindians of lowland Bolivia, the results of which have been reported in the lead author's dissertation and a series of follow-up journal articles. In one paper, they analyzed the impact of three market-related explanatory variables (village distance to market town, cash income, and household wealth) and three acculturation-related explanatory variables (school grade, father's school grade, and fluency in the national language) on folk knowledge of useful plants, which was measured as a function of agreement among informants. The tests of significance carried out for the different variables produced inconsistent, and therefore inconclusive, results regarding the impact of acculturation or market integration on plant use knowledge. Thus, in reference to the proxy variables of acculturation, schooling was positively correlated with higher consensual knowledge while national language fluency and father's schooling was negatively correlated. With respect to the proxy variables of market integration, distance from a market town was associated with higher knowledge but when distance was controlled, neither income nor wealth showed significant correlation. The authors' interpretation of these results is that schooling may foster greater agreement due to common education and greater interaction with each other while market participation may entail different activities, some of which decrease dependency on the forest (e.g. wage labor) while others increase dependency (sale of forest products). In another paper, they compared theoretical ethnobotanical knowledge of plant use (i.e. multiple choice test regarding possible uses of particular species) with ethnobotanical skills (i.e. reports of having manufactured items from plants) and found only a weak association between them and higher variation in regards to skills. They also tested for the effects of participation in wage labor and in sales of farm and forest products on these two classes of ecological knowledge and found that the former activity is associated with fewer ethnobotanical skills (but not lower theoretical knowledge) while the latter activity is associated with greater theoretical knowledge and skills. The authors conclude that some but not all market-related activities (in this case wage labor) may erode local ecological knowledge and that practical skill knowledge or actual use will be more vulnerable. In comparing the general results reported in these two papers, we may well conclude that the results obtained about TEK variation and change depend a great deal on what types of knowledge and what explanatory variables are chosen for study and how they are measured.

5.4. Complex Results and Methodological Variance

The body of research on TEK variation, change and continuity reviewed here has advanced our understanding of the current state and trends of TEK in many parts of the world through the reporting of richly detailed empirical data and systematic analysis of the significance of different environmental variables. A main conclusion that can be drawn from a collective reading of this work is that TEK erosion is a widespread trend but also that it is not universal (and therefore not inevitable). Moreover, where the hypothesis of erosion is supported by the available data, the particulars of this process - i.e. rate of decline, types of knowledge being lost, persons affected, and conditioning factors - vary considerably across groups and sites. While the variability of results strongly suggests the culture- and site-specific nature of this process, it should also be pointed out that this may also be partly due to the different research designs and methods used. The ways that knowledge is defined and measured, the sampling selection, and the variables included in the analysis influence the research outcomes and conclusions. Most individual studies of TEK variation and change are limited to single communities or ethnic groups, or occasionally encompass several communities within the same local area, and are focused on particular, often narrowly defined, knowledge domains to the exclusion of others. There are very few investigations where the same exact method and sampling strategy has been applied to diverse settings and in the few cases where comparative analyses has been attempted it is realized in a post hoc fashion. Clearly the lack of methodological uniformity inhibits comparability and constitutes the main impediment for a more simplified comprehension of the complex empirical reality. It is precisely because of this lack of systematic comparability that we consider that the preexisting data sources cannot be used reliably to develop the TEXVI. Instead, the indicator will require the design of a standardized data instrument and the collection of primary data through its application in different localities.